The requirements of the 5th generation of mobile networks (5G) and the demand for a new 5G air interface are driven by, among others, the handling of massive machine communication (MMC) and the support of low latency (in the order of 1 millisecond) for ultra-reliable communication (URC). In order to fulfil 5G requirements in terms of data rates and latency a new air interface designed to operate at higher frequencies is needed, e.g. above 6 GHz. In comparison to the current frequency bands allocated to LTE, much more challenging propagation conditions exist so that the coverage of the new air interface can be spottier, i.e. more irregular. The extensive use of beamforming, in particular at the network side, may be an essential part of high-frequency wireless access in order to overcome the propagation challenges. Despite potential link budget gains, reliability of a system purely relying on beamforming and operating in higher frequencies might be challenging. In particular, coverage might be more sensitive to time/space variations. Hence, a tight integration of the LTE air interface and the new air 5G air interface, which is referred to as New Radio (NR) below, has been proposed.
With regard to specific architectures to realize a tight integration of LTE and NR (5G), it may be assumed that there are going to be common functionalities in the RAN protocol stack, in contrast to the current interworking between the different accesses such as a common PDCP layer for LTE and NR. In current systems (e.g. UMTS and LTE), differently from that assumption, interworking relies on inter-node interfaces, for both User Plane (UP) and Control Plane (CP). For example, in the case of E-UTRAN and UTRAN interworking, MME and S-GW are inter-connected via the S11 interface. Such an architecture basically allows coverage continuity and load balancing only via hard handovers (always involving core network signalling) and semi-independent resource management for the multiple air interfaces.
A tight integration between LTE and NR has been captured as a requirement in 3GPP TR 38.913, “Study on Scenarios and Requirements for Next Generation Access Technologies”, v0.3.0, http://www.3gpp.org/DynaReport/38913.htm. here, it is noted that the “RAN architecture shall support tight interworking between the new RAT and LTE considering high performing inter-RAT mobility and aggregation of data flows via at least dual connectivity between LTE and new RAT. This shall be supported for both collocated and non-collocated site deployments.” A corresponding objective was captured in the study item on New Radio Access Technology, 3GPP TSG RAN Meeting #71, RP-160671, New SID Proposal: Study on New Radio Access Technology. In order to realize the tight integration of LTE and the new 5G air interface, Da Silva et al., “Tight Integration of LTE and the new 5G air interface to fulfill 5G requirements”, 2015 proposes a logical architecture relying on common RRC/PDCP protocol layers, as shown in FIG. 1A for the control plane. An alternative for common inter-node interfaces (X2*) is further shown in FIG. 1B for the case of standalone LTE and NR network nodes and a co-located LTE/NR network node, i.e. a network node in which LTE and NR are both implemented.
The simplest way to achieve a high performing inter-RAT mobility, in connected mode, between NR and LTE would be to define a single PDCP layer for both LTE and NR, i.e. a single PDPC specification and a single evolution track for both LTE and NR. This single PDCP could be the LTE-PDCP that is enhanced to also rely on services from NR lower layers or a new NR-PDCP specification that has LTE-PDCP as a starting point. In any of these cases, a handover between LTE and NR could possibly benefit from a continued PDCP context and retransmission of PDCP SDUs. In addition to the LTE-PDCP functions associated with the support of high performing mobility, other PDCP functions such as header compression and decompression and in-sequence delivery will also be needed in NR. Therefore, a single PDCP seems to be a reasonable starting point for the UP design to achieve a high performing inter-RAT mobility. This is shown in FIG. 2.
A single PDPC may also be beneficial for multi-RAT aggregation solutions, where PDCP PDU routing for transmission and PDCP PDU reordering for reception functions from LTE-PDCP could be reused. In that case, either NR or LTE may possibly be defined as the RAT where the flow is split. Therefore, the LTE-PDCP may rely on services from the lower layers of NR, as shown in FIG. 3.
With regard to the RRC design to support tight integration of NR and LTE, a particular aspect is related to the RRC support of dual connectivity. Here, for the support of active mode (or Connected state) transmissions for the UE one option would be to have a single set of RRC specification for both LTE and NR. Variants of that option are i) the extensions of LTE RRC functionality to cover NR functions or ii) the creation of a new set of specifications for NR that has LTE functions as starting point. Another option would be to have two RRC specifications, where some level of coordination may be envisioned, e.g. the definitions of some IEs across both specifications.
Regarding the active mode behavior support of dual connectivity, a single or a dual RRC machine may be defined. Here, a single RRC state machine is illustrated in FIG. 4. As such, most of the efforts so far have been concentrated on how to explore the dual connectivity enabled by LTE and NR when the UE is active, i.e. is transmitting some user plane (UP) data or even control messages. Some solutions may be found in Da Silva et al., “Tight Integration of LTE and the new 5G air interface (AI), also referred to as NR here, to fulfill 5G requirements”, 2015.
FIG. 5 summarizes some of the features described in the Da Silva et al. In particular, within Control Plane Diversity a common control plane for LTE and NR would allow a dual-radio UE, i.e. a terminal device having both a dual receiver and transmitter in order to simultaneously connect to both LTE and NR, to have a single control point for dedicated signaling connected via the two air interfaces (AI). The main benefit here is to provide reliability without the need for explicit signaling to switch air an AI, which may be important in certain propagation scenarios where the connection on one AI is lost so quickly that no explicit “switch signaling” could have been performed. Further, in Fast Control Plane Switching the UE would be able to connect to a single control point via any of the AIs and switch very fast from one link to another without the need of an extensive connection setup signaling. This solution may also be used for other UE types than the dual radio UE, i.e. UEs having only a dual receiver but a single transmitter or single radio UEs capable of both AIs but only one at a time. User Plane Aggregation may allow a single data flow to be aggregated over multiple AIs, or to map different data flows on different AIs. In Fast User Plane Switching the user plane for one UE uses only a single AI at a time but a fast switching mechanism is provided between them. Such a mechanism may be applied for all types of UEs. Further, Lean by help of LTE is a feature to let the NR (new AI) transmitter be active when there are active UEs on NR, and to transmit information to idle mode UEs, e.g. system information, over LTE.